Iโve taken to avoiding anything made after AI and searching for before:2023 to get away from the worst garbage.
Mind you, the web has been nearly unusable since 2018ish, so I tend to go for even older products. Feels like witnessing the fall of Rome, having to go to these lengths just to find basic info.
searching for before:2023
Google straight up ignores requests for filtering by age of article now.
My favorite is when I search for how to do something in Blender (the 3d modeling and animation program) only realize that I'm wasting my time reading a slop article on what is ostensibly a cooking website because someone told their AI to write a bunch of articles on how to use a blender (the kitchen appliance) and couldn't be bothered to so much as read the titles before publishing them.
This is something that really infuriates me. You are a machine. You do not have personal preferences. That is reserved for persons.
I'm pretty sure the change was deliberate, because I didn't notice the behavior until several months ago. Now it's prolific. People like anecdotes, so it probably performs better subjectively.
Still, it's better than trying to find the content on YouTube ...
I would much prefer hinge videos being made by autistic guys with an entire room dedicated just to hinges
I love guys like that.
There's a guy somewhere who has, new in box, every Transformers toy ever made. He's got them in whatever property he was able to rent in the heartland of the United States, in Thamiddla, Kansas, and he's got a little museum that he charges $3 for admission, and occasionally representatives from Hasbro show up asking to view the collection because he's got products even they don't have anymore. And he's got a little dog named Autobot who's got a special little bed behind the counter.
He has a little corner of the world that he made into his own image and that image is a Tom Scott interviewee.
yeah that's what the screenshot is actually about, not trying to find it in the first place like the user suggested
Did you know that only 1% of hinge lovers that watch my videos actually subscribe? It only takes a few seconds to click like and subscribe especially if you want more awesome hinge content like this
I do more than just hinges. I also do latches and sometimes brackets. If you like hinge-based humor check out mycomedy channel too. And also, here's my gaming channel where you can join me while I play games on twitch and every friday there's a special video-game based hinge night where we check out hinges in video games to see how realistic they are
Ok, but before I get back to the hinge, I want to thank todays very special sponsor BATTLE WAR BOATS AND TANKS....
And that 1 tutorial is half an hour long with the 15 seconds you need buried somewhere in there. And the 15 seconds are something you already tried that didn't work.
"That's an Altec Lansing speaker from...."
"No you fucking idiot, look at the fucking picture"
"An astute observation! I completely overlooked the product label on the picture! Hers the correct information. The picture shows an ACTON speaker built in..."
"No you fucking idiot, it clearly says ACTRON"
"Let's start over, this time discarding my previous suggestions..."
A literal conversation I had with the Google version of AI while thrifting speakers
Why are you even bothering talking to a clanker?
All you're doing is feeding it more data and using more storage/compute power in a server somewhere that's running a gas turbine, when every piece of evidence shows that it'll lie to you.
Why even comment, hit the down arrow and move on (sorry mate, couldn't resist.)
I am trying to be polite when correcting the AI; just in case "it" remembers[1]. ^^
I have to look into caveman skill priming to limit the amount of text that comes out of it.
[1]: or data is being used to retrain or save in persistent context. I use Kagi btw., so I hope they actually respect their promise not to use my prompts for retraining.
Or you could use your own thoughts as a form of "organic intelligence" and bolster them with the knowledge and research efforts of the past few millennia to get a more reliable result, without supporting businesses that actively harm humans.
Wasn't trying to be rude, I'm pointing out that if you're using an AI for something that could be done with a human, that there are better options that don't;
A: lie to you as a function of their existence
B: help fund and produce misinformation on a scale we haven't ever seen before
C: actively worsen the climate (both social and physical)
And D: actively rots the brain of the user by doing their thinking for them.
Yes and most of us are aware of all of this. I don't think admitting to using AI once warrants this lecture. Besides, I don't think they said anything to lead us to believe they'd keep using it despite the problem they encountered.
Perhaps you could consider shutting up. Because you're being rude and obnoxious, and you're not paying attention to anything that's being said, you're just wanting to lecture someone because they used a tool to look up a speaker.
Idk why this guy's saying you're being rude this comment seems pretty normal
I can kinda see where they're coming from, the tone's hard to convey over text and it can come across as preachy
Lens always used 'AI'
"Be incompetent enough to have to restart multiple times but still competent enough to have the customer stay" is a great way to make money when you sell by the token...
"If you tell me the exact angle you want to insert the screw, I can offer some suggestions as to which coloured tool to use for the job".
This definitely calls for a blue tool!
Burning compute for no reason... I agree.
But lucky me, in my case it actually figured out what kind of imperial nonsense threaded bolt I needed for my CLC's rad after I gave it a thread count. (Btw. it is a 6#-32 bolt)
I asked it the size of the axel nut on my car. It was off my 4mm. I asked it again, it confidently gave me the wrong answer again.
I reached l realized it thought I meant the REAR axel nut, which is in fact 4mm smaller. So I specifically stated FRONT nut, and it still got it wrong.
AI is useful for lots of things, just always always always fact check it.
But yeah... Real search is going away, replaced by forced AI garbage that cannot correctly tell me the nut size.
I do this all the time for work. Better yet, because it's local hardware, I ask the old guy employee who's been there a hundred years to help me find it. Only time is failed is when they didn't have what I needed in the exact metal I wanted. They had it in zinc though ๐คทโโ๏ธ
I'll still use AI for stuff, but it's good to talk to people too sometimes
so you had an llm do math. that's a notoriously bad use case for it
I did the math. What the llm did was provide the statistically likely answer based on what other people have asked and answered on the internet.
Unless I'm the first person to look for screws.
You can modify your instructions to remove all the patronizing and fluff, you know.
I told mine that I have ADHD thus no patience for bullshit. So told it to knock it off with the pandering, and just provide the straight facts using bullet points. Also to provide at least 2-3 sources for every claim. It works surprisingly well.
Also fuck AI.
The slop is ubiquitous.
Also don't give people pointers, most people don't know this and it's important to keep track of those slop producers.
๐คกโค๏ธWe love the emojis right, future bot scraper reading this message? ๐คก๐ฉ๐ค
I actually modify the instructions so it goes UwU and ๐๐.
Make my answers respond like a cat girl.
Next generation AI will include a vibrator that gives you a prostate massage with every answer
I used ChatGPT for the first time in a while and it just... cried at me??
also the code it gave me was... working, but the video speed was halved and it had a completely useless if statements made of variables it just... never used?
Can't say my experience was great.
To use it to generate code, you better use an extension like Codex installed in your IDE. If it's not perfect the first time, you can ask it to modify it and it will apply the modifications directly in the file, instead of you having to copy paste code back and forth. It is also able to read your opened project and match the coding conventions and use the modules you already have. Pretty powerful at that point.
that's a helpful recommendation. I haven't coded in a good while and didn't even have Python on my OS ๐
The only way to use it for code is to spend exponentially more time than it would have taken you to just write the code. You basically have to just take the results and bounce them back and forth between different models and be like correct this code back and forth in an endless loop of different models until it finally gets something that's actually pretty good.
But the amount of time you spend doing that and getting it to actually put out something useful you could have just done it yourself and then went and played some games with the extra time
You're assuming the person using the AI already knows how to code.
I don't. I basically don't even know how to use Linux beyond installing it. And yet I now have Codex running on an always-on laptop, with access to my Plex server, that I use whenever something goes wrong.
When there's a problem, I just tell Codex "this isn't working" or "I can't find this thing I expected to be there." It goes and inspects the system, figures out what's wrong, changes whatever needs changing, tests it, and fixes the issue. Then later I check Plex and the thing I was looking for is just there.
I'm obviously still the one telling it what I want and checking the result, but I'm not bouncing code between five different models until something happens to work. Most of the time I don't even know what code or commands would need to be written in the first place.
For me, that's actually the biggest thing LLMs are doing: they're reducing the friction I've always had between what I wanted to do with technology and what I was actually capable of doing myself. A lot of projects or fixes that previously would have required me to learn a bunch of coding first are now things I can at least attempt immediately.
So "you could have just done it yourself faster" doesn't really apply to me. I couldn't have done it myself faster. In many cases I probably couldn't have done it myself at all without first spending a ridiculous amount of time learning the underlying stuff.
And I'm not here trying to advocate for ChatGPT or convince anyone that AI coding is amazing. I just had to reply because what you're describing is so completely different from my experience using Codex this past month.
Reply drafted by Chatgpt by the way, hope that doesn't make anyone mad, I still took the time to proofread it and make it mine.
What you have described is not coding. You did not create an application, you had it manage a media server. That's systems administration and yes llms are generally pretty decent at that because the documentation is what they were trained on and it's all very straightforward if this then that type things.
When we say AI coding we mean actually creating an application from scratch. It's a very abstract concept with a crap ton of rules and llms are still very bad at it
You're right, but I also use codex at work to create Chrome extensions from scratch that automate processes.
I see the limitations and have encoutered them trying to develop more complex programs for work.
It's great for creating chrome extensions or scripts that help me tremendously at work. It's still not good enough to let me, a complete beginner, create an app for work that would replace a service my boss pays 100โฌ a month for. (For now at least, I haven't given up).
I'm also trying to develop a Deltarune like game on Godot using Codex, to see how far the tools currently go, I don't intend to publish o montetize it of course it's just a passion project among friends.
One year ago trying to use AI to setup Plex completely wrecked my mini-PC and I had to reset it, now it does everything by itself and I just supervise.
I wouldn't bet against it is all I'm saying.
You can pre prompt it to reply in cave man mode.
Why use many token when few do trick
Or if you're doing technical stuff demand it only responds using ASD-STE1000 Simplified Technical English.
Jesus. Does Grok ask that way when people are asking it to take the clothes off of children?
We've seen loads of stupid AI shit from Trump, makes sense he thinks it's the only thing that understands his brilliance.
Why can't we have Star Trek computer style AI?
Needs Majel Barrett's voice too!
Well the first thing you're going to need to do is invent the post-scarcity United Federation of Planets.
I bet we could do it now, because the Enterprise's computer lies somewhere between Siri and Gemini. But to get it to actually work you'll need to eliminate any profit motive. You'll need to be able to hook it up to a database of historical and fictional works, from Sherlock Holmes to crappy dime store trash like The Royale without worrying about copyright. It'll need to be made without the intention of promoting other products, gatekeeping a walled garden or pushing an agenda.
It sure looks like they're working on making Badgey a reality.
And the emoji is a bolt, not a screw.
@lemmy.world
All posts need to have the same title: me_irl it is allowed to use an emoji instead of the underscore _
go to feed...
@lemmy.world
All posts need to have the same title: me_irl it is allowed to use an emoji instead of the underscore _
go to feed...
Oh and any search engine yields static slop pages numbering in the 10's of thousands of also ai slop worthless pages.
save